[SOLVED] Model F AT Soarer Teensy wiring problem

[fohat.digs]

I am trying to add an internal Teensy to a Model F AT and am getting a lot of spurious signals.

The unit works, because Aqua Key Test shows that keys are registering, but there is a great deal of additional noise and lots of keys firing without being touched.

I suspect that it is a wiring or grounding problem. The stock AT cable has a ground wire that connects to the case and the secondary controller or whatever the thing is under the LEDs (attached with the yellow cable).

The Teensy has a "Ground" connection that takes one of the 4 basic wires, but is this cable ground a different circuit?

In this photo from YakMN's guide, I see a 2nd wire grounded to the post, but it is too dark to see what it is coming from.

Could I connect this to the ground on the Teensy? I can't figure out anywhere else to bring it in from.

Thanks

[njbair]

I am trying to add an internal Teensy to a Model F AT and am getting a lot of spurious signals.

The unit works, because Aqua Key Test shows that keys are registering, but there is a great deal of additional noise and lots of keys firing without being touched.

I suspect that it is a wiring or grounding problem. The stock AT cable has a ground wire that connects to the case and the secondary controller or whatever the thing is under the LEDs (attached with the yellow cable).

The Teensy has a "Ground" connection that takes one of the 4 basic wires, but is this cable ground a different circuit?

In this photo from YakMN's guide, I see a 2nd wire grounded to the post, but it is too dark to see what it is coming from.

Could I connect this to the ground on the Teensy? I can't figure out anywhere else to bring it in from.

Thanks

You always want ground to be the same for everything. But that doesn't sound like your problem.

Could it be a faulty main controller? Do you have a known-good spare that you could swap in its place?

[Parak]

I don't recall which pin it is on the AT, but you should be using one of the pins of the controller that the original cable used, not the case ground. White wire on the original cable, IIRC.

[fohat.digs]

Thanks for the responses.

Yes, the white wire is the ground, but the additional ground in the original cable that connects to the screw post and is probably from a sheath or something.

The keyboard worked properly before I cleaned it and started rewiring. I suppose that I will put it back to stock and plug the original cable into an outboard Teensy box. I have never bothered with this on an AT before, since it is plug-and-play with an AT-to-PS/2 adapter, but I need the Teensy now to accommodate the key swap in the ANSI mod.

[Parak]

Yeah, on the AT I have at work, with an internal teensy, I haven't had to connect the teensy to the case. Just the 4 wires to the controller. I'd suspect maybe a wonky connection between the original controller and plate ground?

[fohat.digs]

It turns out that there was no wiring problem at all.

I had cut a new mat for the AT from the dense 1/16" (1.5mm) art foam that I like to use, and it was simply too thick to allow the internal plates to seat together properly. With my F-122 mods, I have added bolts (only 3, these days, but that is enough) down the center of curvature to pull the plates together tightly. I think that the smaller boards, with no center tabs or devices to hold them tight, simply bow out too far and make the center rows too unstable.

Rather than drill and add bolts (in the tight layout there aren't good appropriate positions anyway) I re-cut a new mat from a much thinner material (a reinforced foam used for packaging) and the plates went together nicely and the random/firing/ghosting problem was solved.

I had forgotten how different these small Model Fs really are. The F-122 aircraft carriers really are unlike anything else.

[njbair]

The lesson here is to test early and often. Even if you're doing something you've done a thousand times!

[fohat.digs]

The lesson here is to test early and often. Even if you're doing something you've done a thousand times!

Well, this was the test. The big problem with Model M/Fs is that they have to be completely assembled to test, and completely disassembled to change. I just assume that it is going to take 3 iterations every time. Imagining that it will work properly first time is just dreaming.

I'm glad that I recognized that the problem was probably mechanical rather than electrical.

When I re-connected the original cable and saw the same behavior, I knew that the Teensy was not at fault.

[njbair]

The lesson here is to test early and often. Even if you're doing something you've done a thousand times!

Well, this was the test. The big problem with Model M/Fs is that they have to be completely assembled to test, and completely disassembled to change. I just assume that it is going to take 3 iterations every time. Imagining that it will work properly first time is just dreaming.

I'm glad that I recognized that the problem was probably mechanical rather than electrical.

When I re-connected the original cable and saw the same behavior, I knew that the Teensy was not at fault.

I guess I was thinking to do the cable swap first, test it, then do the mechanical reconditioning and test again. But it sounds like you pretty quickly thought to double-check that anyway, so no big deal.